Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "Identification of Bright Lenses from the Astrometric Observations of Gravitational Microlensing Events"

See other formats


Mon. Not, 



R. Astron. Soc. 000, [|-?? (0000) Printed 1 February 2008 (MN 1*1^ style file vl.4) 



Identification of Bright Lenses from the Astrometric 
Observations of Gravitational Microlensing Events 



Cheongho Han & Youngjin Jeong 

Department of Astronomy & Space Science, Chungbuk National University, Chongju, Korea 361-763 
E-mail: cheongho,jeongyj@astronomy. chungbuk. ac.kr 






> 

in 

o 
(^ 
a^ 

:^ 

Oh' 

o 

U 



X 



Accepted Received 



ABSTRACT 

When a source star is gravitationally microlensed by a dark lens, the centroid of the 
source star image is displaced relative to the position of the unlensed source star with 
an elliptical trajectory. Recently, routine astrometric follow-up measurements of these 
source star image centroid shifts by using high precision interferometers are proposed 
to measure the lens proper motion which can resolve the lens parameter degeneracy 
in the photometrically determined Einstein time scale. When an event is caused by 
a bright lens, on the other hand, the astrometric shift is affected by the light from 
the lens, but one cannot identify the existence of the bright lens from the observed 
trajectory because the resulting trajectory of the bright lens event is also an ellipse. 
As results, lensing parameters determined from the trajectory differ from those of a 
dark lens event, causing wrong identification of lens population. In this paper, we 
show that although the shape and size of the astrometric centroid shift trajectory 
are changed due to the bright lens, the angular speed of centroid shifts around the 
apparent position of the unlensed source star is not affected by the lens brightness. 
Then, one can identify the existence of the bright lens and determine its brightness by 
comparing the lens parameters determined from the 'angular speed curve' with those 
determined from the trajectory of observed centroid shifts. Once the lens brightness 
is determined, one can correct for the lens proper motion. Since the proposed method 
provides both information about the lens brightness (dark or bright) and the corrected 
values of the physical parameters of the lens, one can significantly better constrain 
the nature of MACHOs. 

Key words: gravitational lensing - dark matter - astrometry 



1 INTRODUCTION 

Searches for Massive Astronomical Compact Objects (MA- 
CHOs) by detecting light variations of stars caused by gravi- 
tational microlensing have been performed by several groups 
(MACHO: Alcock et al. 1993, 1996, 1997a, 1997b; EROS; 
Aubourg et al. 1993, Ansari et al. 1996, Renault et al. 1997; 
OGLE; Udalski et al. 1994, 1997; DUO; Alard & Guibert 
1997). With their efforts, more than 300 events have been 
detected toward the Galactic bulge and ~ 20 events toward 
the Magellanic Clouds. 

Despite the large number of the detected events, the in- 
terpretations of the results of microlensing searches toward 
both fields are arguable. Toward the Galactic bulge, the de- 
termined microlensing optical depth of tgb ~ 3.915! * ^ 10~^ 
(Alcock et al. 1997a) substantially exceeds the optical depth 
of r '^ 6 X 10~^ predicted for a standard Galactic disk 
(Paczyiiski 1991; Griest et al. 1991). One explanation for 
the excess optical depth of bulge events would be the pres- 



ence of undiscovered dark populations of disk lenses (Bah- 
call 1984a, 1984b; Alcock et al. 1995). Another possibil- 
ity is that the lenses comprise an ordinary stellar popula- 
tion, and the excess optical depth is due to the existence of 
non-axisymmetric structure in the Galactic bulge (Kiraga 
& Paczyiiski 1994; Zhao, Spergel, & Rich 1995). Toward the 
Magellanic Clouds, on the other hand, the combined mi- 
crolensing optical depth of the MACHO and EROS survey 



of TLMC = 2.1 



+ 1.3 



10"'' (Alcock et al. 1997b; Renault et 



al. 1997) is significantly below the expected optical depth of 
r ~ 4.7 X 10"'' for a halo composed entirely of MACHOs. In 
addition, the time scales of the detected microlensing events 
indicate an average lens mass of ~ 0.5 Mq. The simplest 
interpretation of these results is that nearly half of the halo 
is composed of baryonic matter with a typical mass of order 
of 0.5 Mq such as Hydrogen burning stars and white dwarfs. 
However, both of these lens candidates have difficulties to 
reconcile with other observations (Gould, Bahcall, & Flynn 
1996; Gilmore & Unavane 1999 for the stellar interpretation 



© 0000 RAS 



2 a Han & Y. Jeong 



and Graff & Freese 1996; Adams & Laughlin 1996; Chabrier, 
Segretain, & Mera 1996; Gibson & Mould 1997 for the white 
dwarf interpretation) , leading to alternative scenarios. Sahu 
(1994) and Wu (1994) suggested that a large fraction of 
the Magellanic Cloud events could be due to microlensing 
by stars within the Magellanic Clouds themselves. On the 
other hand, Zhao (1998a; 1998b) suggested that the Mag- 
ellanic Cloud events may be due to lens objects in dwarf 
galaxies or tidal debris from a disrupted galaxy along the 
line of sight to the Magellanic Clouds. 

The ambiguity in the nature of MACHOs arises be- 
cause it is difficult to obtain information about the phys- 
ical parameters of individual lenses from the Einstein time 
scale, which is the only observable from current experiments. 
Detection of microlensing events are accomplished through 
measurements of the variability of source star brightness 
caused by gravitational lensing effect. The light curve of a 
lensing event is represented by 

1/2 



A = 



u^ + 2 



U(u2+ 4)1/2' 



r + 



t-to 



(1.1) 



where u is the lens-source separation in units of the angu- 
lar Einstein ring radius 9e, and the lensing parameters /3, 
to, and fE represent the lens-source impact parameter, the 
time of maximum amplification, and the Einstein ring ra- 
dius crossing time (Einstein time scale), respectively. Once 
the light curve of an event is observed, these lensing param- 
eters are obtained by fitting the observed light curve to the 
theoretical ones in equation (1.1). One can obtain informa- 
tion about individual lenses because the Einstein time scale 
is related to the physical parameters of the lens by 

VE f4GM DoiDi,\^/^ .^ . 

tE=—; TE = I — ^ ;i^ ) , (1.2) 






Dos 



where te = DoiOe is the physical size of the Einstein ring 
radius, v is the lens-source transverse speed, M is the lens 
mass, and Doi, Dis, and Dos are the separations between the 
observer, lens, and source star. However, since the Einstein 
time scale depends on a combination of the lens parameters, 
the values of the lens parameters determined from it suffer 
from large uncertainties. Therefore, to resolve the arguments 
in the interpretation of the result of the lensing experiments, 
a method that can resolve the lens parameter degeneracy for 
general microlensing events is essential. 

A large part of the arguments about the nature of 
gravitational lenses can also be resolved if one can iden- 
tify whether the lens is bright or dark. If a large fraction of 
Galactic bulge events turn out to be caused by dark lenses, 
the most probable scenario for the observed Galactic bulge 
events will be that a significant fraction of Galactic matter 
in the disk is composed of dark lens population. If similar 
results are obtained for Magellanic Cloud events, one can 
exclude the scenario of the self-lensing by stars in the Mag- 
ellanic Clouds themselves. Ideally, one can identify the ex- 
istence of a bright lens because blended light from the lens 
will distort the shape of the microlensing event and shift 
the color of the observed star during the event (Buchalter, 
Kamionkowski, & Rich 1996). In practice, however, it is very 
difficult to detect the presence of a blend by purely photo- 
metric means (Wozniak & Paczyriski 1997). Furthermore, 
since distortion and color change in the light curve can also 
occur due to other types of blend such as nearby unresolved 



blended stars and binary components (Han & Kim 1999), de- 
tection of these symptoms of blending does not always imply 
the detection of a bright lens. As a result, no event has been 
identified as a bright lens event by using this method. 

Recently, routine astrometric follow-up observations of 
microlensing events with high precision interferometers are 
proposed as a method to partially break the lens parameter 
degeneracy of general microlensing events. When a source 
star is gravitationally microlensed by a dark lens, the loca- 
tion of the apparent source star image centroid is displaced 
with respect to the position of the unlensed source star and 
the trajectory of the centroid traces out an ellipse (astromet- 
ric ellipse) during the event (see § 2). Since the size of the 
astrometric ellipse, i.e. the semi-major axis, is directly pro- 
portional to 6e, one can determine the lens proper motion by 
/i = ^e/^e combined with the photometrically determined 
Einstein time scale. While the Einstein time scale depends 
on three lens parameters (M, Doi, and v), the proper motion 
depends only on two parameters (M and Doi)- Therefore, 
by measuring the lens proper motion, the uncertainty of the 
lens parameters can be significantly reduced (Miyamoto & 
Yoshii 1995; H(Zg, Novikov & Polnarev 1995; Walker 1995; 
Miralda-Escude 1996; Paczyriski 1998; Boden, Shao, & Van 
Buren 1998; Han & Chang 1999). When a source star is 
microlensed by a bright lens, on the other hand, the cen- 
troid displacement is distorted by the flux of the lens. How- 
ever, the resulting trajectory of the centroid shifts is also 
an ellipse, and thus one cannot identify the existence of the 
bright lens from the shape of the observed centroid shift 
trajectory (Jeong, Han, & Park 1999). In addition, the de- 
termined lensing parameters from the shape and size of the 
observed centroid shift trajectory will differ from the values 
for the dark lens case, causing wrong determination of lens 
parameters. 

In this paper, we show that although the shape and 
size of the astrometric centroid shift trajectory is changed 
due to a bright lens, the angular speed of the source star 
image centroid shifts around the apparent position of the 
unlensed source star is not affected by the lens brightness. 
Then, one can identify the existence of the bright lens and 
determine its brightness by comparing the lens parameters 
determined from the 'angular speed curve' with those deter- 
mined from the trajectory of observed centroid shifts. Once 
the lens brightness is determined, one can correct for the lens 
proper motion. Since the proposed method provides both in- 
formation about the lens brightness (dark or bright) and the 
corrected values of the physical parameters of the lens, one 
can significantly better constrain the nature of MACHOs. 



2 ASTROMETRIC SHIFTS OF SOURCE STAR 
IMAGE CENTROID 

2.1 Trajectory of Astrometric Shifts 

When a source star is gravitationally microlensed, its im- 
age is split into two. The typical separation between the 
two images for a Galactic event caused by a typical stellar 
mass lens is on the order of a milliarcsecond, which is too 
small for direct observation of individual images. However, 
one can measure the astrometric displacements in the source 
star image light centroid by using several planned high- 
precision interferometers from space-based platform, e.g. the 



© 0000 RAS, MNRAS 000, |-?? 



Identification of Bright Lenses 3 



Space Interferometry Mission (littp;//sini.jpl.nasa.go-v), and 



ground-based large telescopes, e.g. the Keck and the Very 
Large Telescope (Boden, Shao, & Van Buren 1998). 

For a dark lens event, the displacement vector of the 
source star image centroid with respect to the position of 
the unlensed source star is related to the lensing parameters 
by 

^"^ (Tx + Z^y); 



ll2+2 



T = 



t-to 



iE 



2.1.1 



where x and y represent the unit vectors toward the di- 
rections which are parallel and normal to the lens-source 
transverse motion, respectively. If we let x — 56c,x and 



y = 

by 



se. 



-b; 6 = /3fe/2(/3 -1-2), the coordinates are related 



2 

a ; 



where 



and 



fe 



2(/32 + 2)1/2^ 



/3 



a (/32 + 2)1/2 ■ 



(2.1.2) 



(2.1.3) 



(2.1.4) 



Therefore, during the event the apparent source star image 
centroid traces out an ellipse (astrometric ellipse) with a 
semi-major axis a and an axis ratio q. Once the astrometric 
ellipse is measured, one can determine the lensing parame- 
ters 13 and &e because the axis ratio is related to the impact 
parameter and the semi-major axis is directly proportional 
to the angular Einstein ring radius. 

If an event is caused by a bright lens, however, the cen- 
troid displacement is affected by the light from the lens, 
causing the observed centroid shifts differ from those for 
a dark lens event. The bright lens affects the observed cen- 
troid shifts in two ways. First, due to the flux of the lens, the 
apparent source star image centroid is shifted additionally 
toward the lens. In addition, the reference position for the 
astrometric centroid shift measurement is not the position 
of the unlensed source star but the center of light between 
the unlensed source star and the bright lens. By considering 
these two effects of the bright lens, the resulting centroid 
shifts of a bright lens event becomes 



V- 



Oe 



u^ 



-(Tx + /?y), 



(2.1.5) 



'j,2 + 2' 

where the deformation factor is related to the flux ratio be- 
tween the lens and the unlensed source star £l/£s by 



V = 



1 + {£l/£s) + (Il/Is) [(m^ + 2) - «(m2 + 4) 



1/21 



[1 + (eL/es)] [1 + {eL/esMu^ + 4y/y{u^ + 2)] . 

(2.1.6) 
In Figure 1, we present some example trajectories of 
source star image centroid shifts caused by bright lenses 
with various lens/source flux ratios and they are compared 
to the trajectory for a dark lens event. From the figure, one 
finds that the trajectory for the bright lens event is also an 
ellipse (observed astrometric ellipse). Since both dark and 
bright lens events result in the same elliptical trajectories, 
one cannot identify the existence of a bright lens just from 
the shape of the observed astrometric ellipse. One also finds 

© 0000 RAS, MNRAS 000, §-?? 





1 i 1 1 1 


1 


1 1 1 1 

dark lens 


1 1 


0.6 














i,/i^- 1.0 








. .. i,/], = 2.0 




0.4 


1 ^^^ 




- 




. iV 


/I 4 

/ / / 

j; / / / 


- 




^X \\r--^ 




^rw'^'^ y 




0.2 


- >>V\\-'^ 


r-i^JY-;^ 


\ ~ 




- (H>X \ 




y/y-- 


] - 






\ \ 
\ \ 




J - 


u 


~ 




~ 




1 1 1 1 1 1 


1 1 1 1 1 1 


1 1 



-0.4 



-0.2 



0.2 



0.4 



,. (0E 



Figure 1. Trajectories of source star image centroid shifts of 
gravitational microlensing events caused by bright lenses with 
various flux ratios between the lens and unlensed source star, 
ihlt-s- Each trajectory is for an example event with an impact 
parameter /3 = 0.7. We mark the position of the source star im- 
age centroid ('x') at several moments during the time interval 
— l.OtE "^t < l.OtE- The arrow extending (with a thin solid line) 
from the origin represents the orientation vector of the centroid 
at each moment with respect to the apparent position of the un- 
lensed source star at the origin. The direction of centroid shift 
motion is marked by an arrow (with a thick solid line). 



that as the lens/source fiux ratio increases, the shape of 
the observed astrometric ellipse becomes rounder and the 
size (measured by the size of the semi-major axis) becomes 
smaller. Since the lensing parameters /? and 0e are deter- 
mined from the shape and size of the observed astrometric 
ellipse, the determined lensing parameters for a bright lens 
event will differ from the parameters for the dark lens event 
(Jeong et al. 1999). 

2.2 Angular Speed Curve 

The motion of the source star image centroid is additionally 
characterized by its angular speed, oj, around the apparent 
position of the unlensed source star. The angular speed of 
the centroid for a dark lens event is determined by 

""- _ a^ar _ /3tE 
dt 



u;{t) = 



dT dt {t 



toY + f5Hl' 



(2.2.1) 



where (j) = Tan~^(/3/T) is the orientation angle of the source 
star image centroid around the position of the unlensed 
source star. In Figure 2, we present the changes in uj as 
a function of time (the angular speed curve) for events with 
various values of the Einstein time scale and the impact pa- 
rameter. From the figure, one finds that the angular speed 
increase as the source star approaches the lens and peaks 



4 C. Han & Y. Jeong 



>^ 4 






1 1 1 1 1 1 1 
13 = 0.5 


1 1 1 


1 1 1 1 1 1 1 1 


fj-P.n rlavs 




- 


- t, = 30 days 




- 


t,,-50 days 






_ tj,= 100 days 




_ 


- 


/ / ~ \ 1 


t 


,f 




V 


- -'-ni^^' 


1 1 


^\>2i£=:^;^_ - 


~ f 1 1 1 1 1 1 1 1 


1 1 1 1 1 1 1 1 1 ~ 







- 




1 




1 


1 1 


1 1 1 1 1 1 1 1 1 1 


- 




6 


: 


t,- 


= 30 


d 


ays 
|S = 0,3 




A 


^ 


to 
-a 


4 


- 








|S = 0,5 
13 = 0.7 




W 


- 


3 


2 



- 






13 = 0.9 


1 1 




- 




r 


1 




1 1 


1 1 1 1 1 1 1 1 1 1 1 



■100 



-50 





t (days) 



50 



100 



Figure 2. Angular speed curves of the source star image centroid 
sliifts for events with various values of the Einstein time scale tg 
and the impact parameter jB. For all events, we adopt io = 0- 




-2 

6m=m^-m^ (mag) 



Figure 3. Changes in the difference between the impact param- 
eters determined from the observed astrometric ellipse, /3, and the 
angular speed curve, /3o, as a function of the magnitude difference 
between the lens and source star, 5m. To better show SjS = 13 — f^g 
in the region of small Sm, we expand the region and present in a 
separate box. 



when t = to- In addition, the peak angular speed increases 
with decreasing Einstein time scale and impact parameter. 
Since the angular speed is uniquely characterized by /3 and 
tE, these lensing parameters can be determined from the 
measured angular speed curve. We note that among these 
lensing parameters the impact parameter can also be deter- 
mined from the shape of the astrometric ellipse (see equation 
[2.1.4]). 

Another interesting finding about the motion of source 
star image centroid is that its angular speed is not affected 
by the bright lens. When a source star is microlensed, the 
source star images with respect to the position of the un- 
lensed source star lies at the same azimuth as the lens, i.e. 
lens-images are aligned along a straight line. As a result, al- 
though the source star image centroid will experience extra 
amount of shift toward the bright lens along the straight line 
connecting the lens and the source, the orientation angle (j>, 
and thus the angular speed curve, will not be affected by 
the bright lens. Since the angular speed curve for the bright 
lens event is same as that of the dark lens event, the lensing 
parameters for the bright lens event determined from the 
angular speed curve, /3o and fs, will be the same as those 
for the dark lens event. In Figure 1, we mark the positions of 
the source star image centroid ('x') at several moments dur- 
ing the time interval — l.OfE < t < l.OtE for events caused 
by bright lenses with various lens/source flux ratios. We also 
mark the orientation vector of the centroid (an arrow with 



a thin solid line extending from the apparent position of the 
unlensed source star at the origin) at each moment. One 
finds that regardless of the lens brightness, the orientation 
vectors at a same moment coincide. 

Since the angular speed curve is not affected by the 
bright lens, while the trajectory of centroid shifts is affected 
by the lens, the value of the impact parameter of a bright 
lens event determined from the angular speed curve, /3o, will 
differ from the value determined from the observed astro- 
metric ellipse, /3. Therefore, by comparing the impact pa- 
rameters determined in two different ways, one can identify 
the existence of the bright lens and determine its flux. Once 
the lens/source flux ratio is determined, one can also cor- 
rect for the value of the lens proper motion. In Figure 3, 
we present the changes in the impact parameter difference 
5/3 = (3 — l3o as a function of the magnitude difference be- 
tween the lens and source star, 5m — rriL — ms- To better 
show the difference 5/3 in the region of small 5m, we expand 
the region and present in a separate box. From the flgure, 
one finds that although 5f3 have different dependencies on 
5m for different values of /3o, the value of 513 is substan- 
tial even for events affected by lenses with low brightnesses. 
For example, the expected impact parameter difference for 
a bright lens event caused by a lens 2 mag fainter than the 
source star is (5/3 ~ 0.1, which is well beyond the uncertainty 
of the astrometrically determined impact parameter (Boden 
et al. 1998). 



© 0000 RAS, MNRAS 000, 0-?? 



Identification of Bright Lenses 5 



3 SUMMARY 

We study the motion of the astrometric shifts of the source 
star image centroid caused by gravitational microlensing. 
Findings from the study and the apphcations of the findings 
are summarized as follows. 

(i) The angular speed of the source star image centroid 
increases as the source approaches the lens and peaks when 
t = to. In addition, the peak angular speed increases with 
decreasing Einstein time scale and impact parameter. Since 
the angular speed of an event is uniquely characterized by 
the lensing parameters /3 and fE, these parameters can be 
determined from the measured angular speed curve. 

(ii) While the observed trajectory of the source star im- 
age centroid shifts is changed by the bright lens, the angular 
speed curve is not affected by the lens. Therefore, although 
the lensing parameters determined from the observed cen- 
troid shift trajectory, i.e. /3 and 6e, differ from those for 
a dark lens event, the lensing parameters /3o and t-E deter- 
mined from the angular speed curve are same as those of the 
dark lens event. 

(iii) One can identify the existence of the bright lens and 
determine its brightness by comparing the impact param- 
eters determined from the astrometric shift trajectory and 
the angular speed curve of the source star image centroid. 
The difference in the two impact parameters is substantial 
even for events affected by lenses with low brightnesses, en- 
abling one to easily identify the lens. 



Walker, M. A. 1995, ApJ, 453, 37 

Wozniak, P., & Paczynski, B. 1997, ApJ, 487, 55 

Wu, X. P. 1994, ApJ, 435, 66 

Zhao, H. 1998a, MNRAS, 294, 139 

Zhao, H. 1998b, ApJ, 500, 149 

Zhao, H., Spergel, D. N., & Rich, R. M. 1995, ApJ, 440, L13 



REFERENCES 

Adams, F., & Laughlin, G. 1996, ApJ, 468, 586 

Alard, C, & Guibert, J. 1997, A&A, 326, 1 

Alcock, C., et al. 1993, Nature, 365, 621 

Alcock, C., et al. 1995 ApJ, 445, 133 

Alcock, C., et al. 1996, ApJ, 461, 84 

Alcock, C., et al. 1997a, ApJ, 479, 119 

Alcock, C., et al. 1997b, ApJ, 486, 697 

Ansari, R., et al. 1996, A&A, 314, 94 

Aubourg, E., et al. 1993, Nature, 365, 623 

Bahcall, J. N. 1984a, ApJ, 276, 169 

Bahcall, J. N. 1984b, ApJ, 287, 926 

Boden, A. F., Shao, M., & Van Buren, D. 1998, ApJ, 502, 538 

Buchalter, A., Kamionkowski, M., & Rich, R. M. 1996, ApJ, 469, 

676 
Chabrier, G., Segretain, L., & Mera, D. 1996, ApJ, 468, L21 
Gibson, B., & Mould, J. 1997, ApJ, 482, 98 
Gilmore, G., & Unavane, M. 1999, MNRAS, in press 
Gould, A., Bahcall, J. N., & Flynn, C. 1996, ApJ, 465, 759 
Graff, D, & Freese, K. 1996, ApJ, 456, L49 
Griest, K., et al. 1991, ApJ, 387, 181 
Han, C., & Chang, K. 1999, MNRAS, 304, 845 
Han, C., & Kim, T.-W. 1999, MNRAS, 305, 795 
H!%, E., Novikov, I. D., & Polnarov, A. G. 1995, A&A, 294, 287 
Jeong, Y., Han, C., & Park, S.-H. 1999, ApJ, 511, 569 
Kiraga, M., & Paczynski, B. 1994, ApJ, 430, LlOl 
Miralda-Escude, J. 1996, ApJ, 470, L113 
Miyamoto, M., & Yoshii, Y. 1995, AJ, 110, 1427 
Paczynski, B. 1991, ApJ, 419, 648 
Paczynski, B. 1998, ApJ, 404, L23 
Renault, C., et al. 1997, A&A, 324, 69 
Sahu, K. C. 1994, Nature, 370, 275 
Udalski, A., et al. 1994, Acta Astron., 44, 227 
Udalski, A., et al. 1997, Acta Astron., 47, 169 



© 0000 RAS, MNRAS 000, §-?